Case Study: How a Copper SX-EW Mine Switched Acid Sources Without Stopping Production

In copper solvent extraction–electrowinning (SX-EW) circuits, sulfuric acid serves simultaneously as the primary leaching reagent and the dominant electrolyte component. Oxide and secondary sulfide ore heaps are irrigated with raffinate acidified to 5–15 g/L H₂SO₄, producing pregnant leach solution containing 2–5 g/L Cu at pH 1.5–2.0. The same acid supply network maintains tankhouse electrolyte at 160–200 g/L H₂SO₄ with copper cathodes plated at current densities between 250–350 A/m². Acid consumption for oxide copper leaching commonly ranges from 2.5–3.5 t H₂SO₄ per tonne of copper cathode where acid-consuming gangue is present. Because the acid feed is tied directly to both heap leaching and metal recovery, a change in acid source produces simultaneous shifts in leach pH, solvent extraction chemistry, electrolyte impurity load, and tankhouse operating stability.

Concentrated sulfuric acid supplied to SX-EW operations is received at 93–98 wt% H₂SO₄ with density between 1.820–1.840 g/cm³ at 20 °C. The product may be produced from sulfur burning, metallurgical off-gas, or spent acid regeneration. Each source carries a distinct impurity profile, and that profile determines the maximum allowable switching rate in an operating circuit.

Physical and Chemical Boundaries for SX-EW Acid Service

Feed acid quality in an SX-EW circuit is controlled less by the overall H₂SO₄ concentration than by the concentration of chloride, nitrate, iron, arsenic, and solid residue. A 98.5 wt% acid from a sulfur-burning plant may carry chloride below 5 mg/kg, while smelter off-gas acid can arrive with chloride between 10 mg/kg and 50 mg/kg depending on gas cleaning performance. This difference, when multiplied by daily acid consumption of 80–120 t/d, introduces a measurable step change in the aqueous chloride load entering the solvent extraction and electrowinning circuit. The impurity transfer to electrolyte is not linear: chloride volatilizes at anode surfaces, participates in pitting of 316L cathode blanks, and accelerates corrosion of Pb-Ca-Sn anodes. Ferric iron and arsenic in acid also partition into the leach solution and can transfer through solvent extraction if pH control is lost.

ASTM E223-16 provides the standardized test framework for sulfuric acid concentration, residue, and common impurities. Total acidity is also determined by titration under ISO 910, while sulfur dioxide is measured by iodometric titration under ISO 3423. In a source-switching program, the receiving laboratory must run full impurity panels at the unloading skid before a tank is released to service. A representative quality envelope for two SX-EW feed sources is shown below.

The critical operational boundary in an SX-EW switch is not the acid concentration itself but the rate at which the impurity inventory of the recirculating aqueous phase changes. Sulfuric acid dilution before heap irrigation reduces impurity concentration immediately, but the retention time of process solutions in a heap stack can exceed 30 days, so disturbances appear in the pregnant leach solution over a similar delay.

Receipt of acid from a second source was integrated into the existing tank farm through a dedicated unloading skid with 316L and EPDM-lined flexible connections. Design flow capacity of 30 m³/h against a discharge head of 25 m allowed a 25 t tanker to unload in approximately 50 min, while normal plant consumption of 80–120 t/d required buffer inventory of at least 7 days. The original sulfur-burned acid tank and the smelter off-gas acid tank were connected through a common stainless steel ring main with double block-and-bleed isolation valves to prevent cross-contamination during switchover. Coriolis mass flow meters with accuracy of ±0.2% of rate were installed on both feed legs, and the control system was configured to reject a source if the differential pressure across the final filtration skid exceeded 2.5 bar.

Can Sulfuric Acid Impurity Step Changes Destabilize Solvent Extraction?

The solvent extraction stage is the most sensitive point in the SX-EW sequence during an acid source change. Hydroxyoxime extractants in a high-flash diluent are operated at organic-to-aqueous phase ratios between 1.0 and 1.5, with mixer-settler specific throughput normally in the range of 4–6 m³/m²·h. A step increase in chloride entering the aqueous feed can alter interfacial tension, promote organic entrainment, and increase crud formation. Field experience from mixer-settlers processing agitated leach solutions indicates that chloride concentrations above 20–30 mg/L in the aqueous feed can measurably reduce phase disengagement and require additional organic recovery capacity. In the switch described here, the transition from low-chloride sulfur-burned acid to smelter off-gas acid was therefore staged over 72 h with the volumetric blend ratio changed in 25 percentage-point increments at 24 h intervals. This staged introduction kept the aqueous chloride concentration at the solvent extraction feed below 20 mg/L throughout the changeover, as confirmed by daily potentiometric titration and ion chromatography.

Iron and arsenic are less likely than chloride to cause immediate solvent extraction failure, but they affect tankhouse chemistry. Arsenic entering the electrolyte at concentrations above 5–10 mg/L can co-deposit or alter cathode nodulation depending on antimony and bismuth interactions. Antimony and bismuth are particularly harmful in the tankhouse at concentrations above 0.5 mg/L, and their ratios to arsenic are monitored more closely than arsenic alone. Because the smelter off-gas acid in the switch contained measurable arsenic, the electrolyte bleed rate was increased from 1.5% to 2.5% of total tankhouse volume per day for the first 14 days after blend completion. This prevented arsenic accumulation above the action limit of 10 mg/L in circulating electrolyte while maintaining electrolyte copper at 35–40 g/L.

When Acid Source Switching Requires Blending Down 98% Acid to Raffinate Addition Points

The physical act of introducing acid from a second source into an existing raffinate addition system is governed by heat of dilution, material compatibility, and pH control stability. Concentrated acid is not injected directly into heap irrigation headers without first mixing with raffinate. In the switch, 98 wt% acid from the new source was metered into a 316L static mixer with a minimum line velocity of 0.5 m/s, discharging into a reinforced polypropylene-lined distribution manifold. Inline pH analyzers with temperature compensation controlled the acid addition valve to maintain raffinate pH between 1.6 and 1.8 at the heap distribution ring. The control dead time from acid injection to the pH sensor was less than 45 s, preventing large pH swings when the acid source was toggled between tanks.

During the 72 h blending window, acid strength at the addition point was held at 96–98 wt% by controlling the ratio of the two feed tanks. The blend ratio was supervised by a mass balance calculation using tank level radar gauges with ±1 mm resolution and Coriolis mass flow totals. If the density-derived H₂SO₄ concentration of the blended stream deviated by more than ±0.5 wt% from the target, the control system automatically returned to the original source until the alarm was cleared. This interlock prevented the addition of off-spec acid to the leach circuit. Published data for this specific switch configuration is limited; the operating ranges above are compiled from standard SX-EW design practice and acid supply specifications rather than a single publicly named mine.

Within the tankhouse, electrolyte composition changed from 165 g/L H₂SO₄ and 38 g/L Cu to 175 g/L H₂SO₄ and 36 g/L Cu over 14 days after blend completion. This drift reflected the slightly higher sulfate-to-acid ratio in the incoming smelter off-gas acid and the increased bleed rate. Current density was maintained at 280–320 A/m² with cell voltage between 1.9 and 2.3 V. Cathode copper from the first two harvests following the switch was inspected for nodulation and edge effects; no cathode pull was required because the organic entrainment in electrolyte remained below 5 ppm v/v as measured by FTIR extraction.

Electrolyte Composition Drift After a Source Switch

Electrolyte monitoring after the acid source change was based on daily sampling of iron, chloride, arsenic, manganese, and copper by ICP-OES with method detection limits of 0.1 mg/L for chloride and 0.05 mg/L for arsenic. The sulfate-to-chloride ratio was maintained above 100:1 to limit anode attack. A chloride spike to 25 mg/L was observed on day 6, corresponding to a delayed release of chloride from the heap inventory. The response was to return the acid blend to 50% smelter off-gas acid for 48 h and increase tankhouse bleed to 3.0% per day. Chloride returned to below 15 mg/L by day 9, and the bleed was reduced to 2.0% per day by day 12. No interruption in leach irrigation, solvent extraction throughput, or cathode harvesting occurred during this period.

Antimony and bismuth were below the limit of quantitation in the new acid source by ICP-MS using a 0.01 mg/L reporting limit, so no additional cobalt addition was required for anode passivation. Tankhouse cobalt concentration was maintained at 100–150 mg/L per standard Pb-Ca-Sn anode practice. The switch demonstrated that source substitution is operationally feasible when the impurity load is controlled at the acid receipt point, staged through the solvent extraction feed, and buffered by adequate tankage and analytical response. The remaining operational constraint was not acid quality but the ability of the heap inventory to damp the chloride transient, which required the extended blend schedule described above.